Sistem Rekomendasi Produk UMKM Menggunakan Algoritma User-Based Collaborative Filtering Berbasis Website

Abstract

Kolaka Regency is one of the regions in Southeast Sulawesi Province that is rich in quality MSME products, but faces obstacles in promoting these products to a wider market. This research aims to adapt and optimise the UBCF method for the Kolaka MSME context and improve the algorithm to overcome the challenges faced, especially data limitations and variations in customer preferences that change quickly. This research method uses a recommendation engine approach with the UBCF method, which is applied through the stages of data preparation, UBCF implementation on a web-based system, and recommendation accuracy testing. The data used is product rating data from users, which is then processed using UBCF. The test results show that this recommendation system is able to provide fairly accurate rating predictions. Based on the results of testing the accuracy of the recommendations that have been carried out, it is obtained for the MAE value of 1.11, then the MSE value of 0.0649 and for the MAPE value obtained of 1.65%. This research contributes to improving the competitiveness of MSME products in Kolaka through UBCF technology, and provides a model that can be applied in other areas with similar characteristics.